Active treatments outperform minimal intervention for adults with rotator cuff tendinopathy: a systematic review with predictive and network meta-analyses of complex interventions

This systematic review and network meta-analysis of 140 high-quality studies involving over 10,000 adults with rotator cuff tendinopathy concludes that active treatments, particularly those focusing on strengthening, range-of-motion exercises, and movement pattern retraining, are consistently superior to minimal intervention across all timeframes, indicating that a "wait-and-see" approach should be avoided.

The Gist

Imagine your shoulder is like a complex, high-performance engine. Sometimes, the parts inside (the tendons) get frayed, irritated, or "tendinopathic" (a fancy word for being angry and inflamed). This condition, called Rotator Cuff Tendinopathy, is incredibly common and causes a lot of pain, making it hard to lift your arm or sleep at night.

For years, doctors and patients have been arguing over the best way to fix this engine. Some say, "Just wait it out; it will heal on its own." Others say, "Inject it with steroids!" or "Do these specific exercises!" or "Get laser therapy!"

This new study, published by a team of researchers, acted like a super-detective. They didn't just look at one clue; they gathered evidence from 140 high-quality studies involving over 10,000 people. They wanted to answer two big questions:

1. Is doing something actually better than doing nothing?
2. If we do something, which specific "tool" works best?

Here is the breakdown of their findings, explained with some simple analogies.

1. The "Wait-and-See" Myth (The Passive Approach)

The Old Idea: Many people thought, "If I just rest my shoulder and wait a few months, the pain will go away on its own."
The Detective's Find: The study found that waiting is a losing strategy.

• The Analogy: Imagine your shoulder pain is a small fire. If you just sit there and watch it (the "wait-and-see" approach), the fire might burn a little smaller for a while, but it won't go out. In fact, after a year, the fire often flares back up again.
• The Result: People who did nothing (or just got a fake treatment) saw only tiny improvements in the short term, and their pain often came back later. The condition is not self-limiting; it needs help.

2. The "Active Treatment" Victory (The Firefighters)

The Finding: Every single active treatment they looked at was better than doing nothing.

• The Analogy: If the shoulder is a fire, active treatments are the firefighters. Whether they used water, foam, or a fire extinguisher, all of them put out the fire better than just watching it burn.
• The Result: Whether it was short-term (3 months), medium-term (6 months), or long-term (1 year), people who got active treatment felt significantly less pain and moved much better than those who did nothing.

3. The "Toolbox" Hierarchy (Which Firefighter is Best?)

This is where it gets interesting. The researchers tried to rank the treatments to see which one was the "Superhero."

• The Problem: It's hard to say one tool is 100% better than another because every patient is different, and studies often mix tools together.
• The Winners: However, a clear "Hall of Fame" emerged. The treatments with the strongest, most reliable evidence were:
  1. Strengthening Exercises: Building the muscle armor around the shoulder.
  2. Range-of-Motion Exercises: Keeping the joint flexible so it doesn't get stiff.
  3. Movement Pattern Retraining: Teaching your brain and body to move correctly so you don't re-injure the shoulder.
  4. Complex Interventions: A "Swiss Army Knife" approach that combines education, exercise, and other therapies all at once.

The Losers (or at least, the less proven ones):

• Injections (Steroids): These are like a "magic potion" that works very fast (like a fire extinguisher that puts out the fire instantly), but the fire often comes back stronger later. The study suggests they are good for short-term relief but not a long-term cure.
• Laser, Ultrasound, and Taping: These are like "flashy gadgets." They might help a little bit in the short term, but there isn't enough proof that they work well over the long haul compared to good old-fashioned exercise.
• Education Alone: Just reading a pamphlet or listening to a lecture is like being told how to put out a fire without actually having a hose. It helps a little, but it's not enough on its own.

4. The "Personalized GPS" Conclusion

The study didn't find one single "magic bullet" that works for everyone. Instead, they found that the best approach is a personalized GPS.

• The Analogy: Think of your shoulder pain as a journey. You need a map (the evidence) and a guide (the doctor/physiotherapist).
• The Advice:
  • Don't wait: Get moving early.
  • Focus on the basics: Prioritize strengthening and moving your shoulder correctly.
  • Mix it up: Combine exercise with learning about your condition (education).
  • Listen to your body: If one exercise hurts, try a different angle. The "best" treatment is the one that fits your life, your preferences, and what your body responds to.

The Bottom Line

If you have a sore shoulder, don't just sit on the couch waiting for it to heal. That won't work. Instead, get active. The most reliable path to recovery is building strength, improving how you move, and combining different therapies under the guidance of a professional. It's not about finding the one perfect pill; it's about building a strong, resilient engine that can handle the road ahead.

Technical Summary

1. Problem Statement

Rotator cuff tendinopathy (RoCuTe) is a prevalent condition causing pain, functional loss, and reduced quality of life (QoL), yet there is no consensus on the optimal conservative management. Previous systematic reviews have been limited by:

• Focusing on single treatments rather than the complex, multi-component interventions common in clinical practice.
• Including low-quality studies, leading to high risk of bias.
• Failing to benchmark active treatments against a data-driven estimate of "minimal intervention" (wait-and-see or sham), leaving uncertainty about whether the condition is self-limiting.
• Lack of clear guidance on treatment hierarchies across short-, medium-, and long-term timeframes.

2. Methodology

This study employed a rigorous Systematic Review combined with Bayesian Predictive Meta-Analysis and Network Meta-Analysis (NMA).

• Data Sources & Selection: A search of PubMed, Embase, Web of Science, CINAHL, and SPORTDiscus was conducted up to August 22, 2025.
  • Inclusion: High-quality Randomized Controlled Trials (RCTs) with a PEDro score ≥7 (plus an added conflict-of-interest criterion).
  • Population: Adults (≥18 years) with clinically diagnosed RoCuTe (traumatic or insidious).
  • Interventions: Any conservative treatment vs. active or minimal intervention.
  • Outcomes: Pain (0–100 scale), Function (Standardised Mean Change, SMC), and QoL (SMC).
• Study Characteristics: 140 high-quality studies were retained, involving 10,260 patients (mean age 48 ± 8 years, 55.9% female).
• Analytical Framework:
  • Benchmarking: Minimal interventions (wait-and-see, inert sham, sham, sham + active adjunct) were pooled to create an empirically derived "minimal-intervention benchmark" for comparison.
  • Predictive Meta-Analysis: Bayesian random-effects models were used to estimate the absolute benefit of active treatments over the minimal-intervention benchmark. The probability of a future study exceeding the Minimal Clinically Important Change (MCIC) was calculated.
  • Network Meta-Analysis (NMA): Integrated direct and indirect evidence to rank active interventions relative to one another (using strengthening as the reference).
  • Evidence Grading: Certainty of evidence was assessed using the GRADE framework, considering risk of bias, inconsistency, imprecision, and publication bias.
  • Stakeholder Involvement: Patient and Public Involvement (PPI) guided the selection of outcomes and research questions.

3. Key Contributions

• Complex Intervention Handling: Unlike previous reviews that discarded complex interventions, this study classified interventions by their primary "treatment focus" (e.g., strengthening, range of motion) while acknowledging adjuncts, allowing for the analysis of real-world clinical complexity.
• Predictive Modeling: The use of Bayesian predictive distributions allows clinicians to estimate the likely outcome of a future trial, rather than just summarizing past data.
• Minimal Intervention Benchmark: The study quantified the natural history and placebo effects of RoCuTe, confirming that the condition is not self-limiting and that even "wait-and-see" approaches yield only modest, often regressive, long-term results.
• Three-Tier Recommendation System: Interventions were categorized based on data volume and certainty of superiority:
  1. Complete evidence, higher certainty.
  2. Incomplete evidence or no higher-certainty comparisons.
  3. Very limited evidence.

4. Key Results

A. Minimal Intervention (The Benchmark)

• Short-term: Small improvements in pain (2.6/100), function (SMC 0.13), and QoL (SMC 0.05).
• Medium-term: Modest improvements in pain (23.3/100) and function (SMC 0.87).
• Long-term: Evidence of regression in pain (21.1/100) and function (SMC 0.76), indicating that without active treatment, symptoms often worsen or fail to sustain improvement.

B. Active vs. Minimal Intervention

All active interventions with sufficient data were superior to minimal intervention at all timepoints.

• Pain Reduction: Active treatments reduced pain by 18.1–45.0 points (0–100 scale) depending on the intervention and timeframe.
• Function Improvement: SMC ranged from 1.0 to 2.4 (Short-term) to 1.0 to 1.8 (Long-term).
• QoL Improvement: SMC ranged from 0.8 to 1.8.

C. Treatment Hierarchy & Recommendations

Based on GRADE certainty and data volume, the following interventions received the strongest recommendation (Group 1: Complete evidence, higher certainty):

1. Strengthening Exercises
2. Range-of-Motion (ROM) Exercises
3. Complex Interventions (multi-component treatments)
4. Movement Pattern Retraining (Motor control/proprioception)

Other Findings:

• Education: Effective for QoL but showed limited efficacy as a sole treatment for pain/function in the short term.
• Injections (Steroids/PRP): Showed short-term superiority but had lower certainty for long-term efficacy and were associated with publication bias.
• Manual Therapy & Shockwave: Showed short-term benefits but lacked long-term data or high-certainty comparisons.
• Network Meta-Analysis: Did not establish a clear hierarchy among active treatments due to wide credible intervals and unstable rankings, reinforcing that multiple active options are viable.

5. Significance and Clinical Implications

• Rejection of "Wait-and-See": The study provides robust evidence that a passive approach is inferior to active management at all stages, even in the short term.
• First-Line Treatment: Clinicians should prioritize strengthening, ROM, and movement pattern retraining, ideally delivered as complex interventions that address multiple patient needs (biomechanical, psychological, and educational).
• Personalized Care: Since no single active treatment is statistically superior to all others in head-to-head comparisons, treatment selection should be guided by clinical reasoning, patient preference, and resource availability.
• Methodological Advance: The study sets a new standard for synthesizing complex interventions in musculoskeletal research, moving beyond simple "single vs. single" comparisons to reflect the reality of clinical practice.

Conclusion: Active management is essential for RoCuTe. The most credible evidence supports a multimodal approach centered on exercise (strengthening and ROM) and movement retraining, rather than passive modalities or observation.